Glenoid lateralization influences active internal rotation after reverse shoulder arthroplasty.

Abstract

Improvement in internal rotation (IR) is not reliably achieved after reverse total shoulder arthroplasty (RSA). The purpose of this study was to evaluate the relationship between postoperative IR and glenoid-sided lateralization following RSA in an implant using metallic lateralization. A multicenter retrospective study of RSAs with minimum 1-year clinical follow-up was performed. Patients were stratified based on the amount of glenoid-sided implant lateralization into 4 groups: 0-2 mm (n = 57), 4 mm (n =238), 6 mm (n = 95), and 8 mm (n = 65). The primary study outcome was active IR at a minimum of 1 year postoperatively, measured both by spinal level and in degrees with the shoulder abducted to 90°. Secondary outcomes were active forward flexion and external rotation, belly press strength, and subjective patient-reported outcome (PRO) measures. Comparisons were made with 1-way analyses of variance. Linear regression analyses evaluating for the association of glenoid lateralization with active IR were also performed to control for additional confounders, including demographics and other implant variables such as glenosphere diameter, humeral lateralization, humeral version, and whether the subscapularis was repaired. A total of 455 patients were included in the study. The mean age was 69 years, and 48% of patients were male. IR differences varied by the method of measurement (spinal level vs. IR in degrees with arm abducted). Overall, patients with 8 mm of glenoid lateralization had significantly improved IR compared with all other lateralization groups. Patients with 6 mm of glenoid lateralization had significantly improved IR compared with the 0-2- and 4-mm groups. There were no significant differences in the secondary outcomes or PROs between lateralization groups. In the regression analysis, glenoid lateralization was the only implant-related variable that was significantly associated with improved IR for both measurement methods. Glenosphere diameter and humeral version were both significantly associated with IR measured in degrees with the arm abducted but not spinal level. For the studied implant system, glenoid lateralization of 6-8 mm was associated with improved active IR at 1 year compared to patients with less glenoid lateralization with no significant differences in active forward flexion, external rotation, or PROs. In a multivariable analysis, increased humeral retroversion was associated with increased IR at 90° and increasing glenosphere diameter was associated with decreased IR at 90°, whereas BMI, subscapularis repair, and humeral lateralization did not significantly affect active IR.